Linux.com

Joe Barr rips proprietary software vendor a new one
Friday March 09, 2007 (09:01 PM GMT)
By: Joe Barr.

And btw, Joe, 90% of the professional programmers are in the proprietary software space. I guess you aren't a professional programmer so you're not qualified to speak for all of us

#

1.) Authors of articles do not always get to pick their own titles, and even if he did, there are times when referring to yourself in 3rd person is appropriate.

2.) I'd like to see you back up that 90% statistic. Maybe a link to a reputable source?

3.) You use your 90% statistic to hypothesize that Joe is not a professional programmer:

      a) Joe never claimed he was one. (Can you show me where he makes this claim?)

      b) Joe is not trying to speak for you... (Again, show me where he attempts to speak for you.)

      c) Even -if- 90% of professional programmers worked in the propriatary world, that would leave 10% who didn't, and Joe could easily fall in that 10%, so I fail to see how you drew that conclusion?

-Jeremy

#

Thanks for explaining article titles to them, Jeremy. I started programming professionally in 1974, and hung up my coding coding pen in, oh, 2001, I think. But as you note, I never claimed to be speaking for programmers at all.

#

So which parking lot are you two guys going to meet to discuss this?

#

I would interpret the following quote - "~95% of code is still written in-house" - as indicating that 5% of code written is actually in development of "for sale" programs. If, of all the code written internationally, only 5% is actually proprietary contributing to a for sale program as the product then you may have got your ration backwards.

Though, read the paper for yourself; gaining more information is never a bad thing. It's from 1999 but I can't see the ratio of inhouse code to for sale programs changing greatly in that time.

<a href="http://www.catb.org/~esr/writings/cathedral-bazaar/magic-cauldron/ar01s03.html" title="catb.org">http://www.catb.org/~esr/writings/cathedral-bazaa<nobr>r<wbr></nobr> /magic-cauldron/ar01s03.html</a catb.org>
(The Magic Cauldron)

"First, code written for sale is only the tip of the programming iceberg. In the pre-microcomputer era it used to be a commonplace that 90% of all the code in the world was written in-house at banks and insurance companies. This is probably no longer the case—other industries are much more software-intensive now, and the finance industry's share of the total must have accordingly dropped—but we'll see shortly that there is empirical evidence that approximately 95% of code is still written in-house."
(ESR 1999)

"This is called `maintenance', and any software engineer or systems analyst will tell you that it makes up the vast majority (more than 75%) of what programmers get paid to do. Accordingly, most programmer-hours are spent (and most programmer salaries are paid for) writing or maintaining in-house code that has no sale value at all—a fact the reader may readily check by examining the listings of programming jobs in any newspaper with a `Help Wanted' section."
(ESR 1999)

#

"First they ignore you, then they laugh at you, then they fight you, then you win." -- Gandhi

#

5 years + is how long it has taken to fix some security holes in windows.

Open Source has the advantage if you think its really important you can get to group of people who think its important as well if a small companys and pay a programmer to fix it. If you are a large company just use your on staff programmers.

Closed Source stuffed until vendor thinks its important. Like after a virus is released go threw that security hole.

Open Source is in your hands. Close Source is in some other company.

#

> But my recollection is that IIS has larger
> market share, and the gap is widening.

totally incorrect. check out the Netcraft surveys. Apache has a much larger share (and always has).

> Why would that be if open source
> were compelling for all users?

it is not.. because your original assertion is not correct.

#

This guy sounds like one more of microsofts mouthpieces who got some lunch money to spew garbage from his word hole...

#

While the assertion that Windows 2000 or 2003 has to be rebooted every other week is pretty much untrue (my experience is that more Windows patches require a restart than linux patches), asserting that because it's proprietary you are guaranteed updates is equally bogus.

Read the software agreements, in detail, on most of your software. It essentially guarantees that you might have the right to run their software, and that's about it.

For the average computer user, or even system administrator, a flaw in the code means you're at the mercy of someone else until a patch is available. Personally, I've found open source projects to be *more* responsive for patches.

By the way, while I could probably run Red Hat 6.0 in a secure fashion (it would be painful to do), Windows 2000 has reached the point where it's not really a good idea to have it publicly accessible any longer. You should upgrade.

#

"...for many applications, it makes more sense for organizations with scarce resources and tight budgets to purchase software from a vendor they hold accountable for keeping the software up to date..."

And that's precisely why I dumped all Windows machines from my business and moved to Linux and Open Office. The vendor of the proprietary products I was using previously not only didn't understand the concept of accountability, but was actively replacing it with arrogance.

So I shelled out a whopping $50 for a Slackware Install set (even though I could have downloaded it for free... but Slackware has been rock solidly accountable for so many years I felt the expenditure was easily justified), downloaded O_O and some other tidbits I needed, tailored the configuration just the way I wanted it, created an image with partimage, and installed it on all 12 of my machines (ouch, what an expenditure of resources... a whole evening of my time and all I did was save probably several thousand dollars over the next few years).

You say this open source stuff may be dangerous to my network, but hey... I guess I just like living dangerously.

#

"...for many applications, it makes more sense for organizations with scarce resources and tight budgets to purchase software from a vendor they hold accountable for keeping the software up to date..."

What is this "accountability" that the bean counters are always yammering about, anyway? The license terms for proprietary software are such that you can't sue them for bugs, up to and including the stuff not working at all or being fit for your intended use. And if the vendor is more than mom-&-pop size, they'll deal with it based on their perception of how important the bug is, and how important you are to them. So any notion of "support" from software vendors is a delusion. And you have no recourse if the product's no good.

Free software ("open source" is a mealy-mouthed term representing an attempt at a political compromise-- let's stick to words that make sense) is of variable quality, depending on the project. In that way it's no different than the commercial stuff. The best have better community responsiveness to bugs than most commercial software firms I've dealt with, and I've dealt with almost all the big 'uns over the decades.

There are niches where proprietary vendors still provide better products. For example I don't think I'd run a semiconductor company's manufacturing systems on free software. But if I were starting a small to medium business, I would have few needs that couldn't be met by free software, besides possibly CRM.


 

#

To clarify only one point for symantic accuracy; Free Software and Open Source are two different groups within F/OSS representing the two major coltural poles (also why it's often written with the slash).

Free Software is specifically the Free Software Foundation and political goals it embodies; all software should be free on principal, closed source is the devil, binary blobs are not exceptable, patents and copywrite should be abolished entirely. While even Richard Stalman says that closed source should be used only when no free alternative is available, this is where you get the extremist Zealots from usually. Free Software is a political goal.

Open Source focuses on the technical advantages of peer review, bug reporting and patch submission; every bug is shallow to someone, source available for in-house security auditing or further development, software evolves quicker when developers share and build off each other's code. Steriotipical examples are Linus Torvalds and Eric S Raymond who present technical arguments to technical problems; example, Linus and the GNOME project where Linus simply wrote and submitted a bunch of patches as his reply to there personal attacks. Further examples are the popularity and stability of Apache, Linux (the kernel), K3B (one of the better disk burning suites).

Free Software and Open Source are two seporate cultures within Free/Libre and Open Source Software. Don't discount the engineers who let there technically stronger software speek for them just because the politically motivated zealots make more noise.

#

Free Software is specifically the Free Software Foundation and political goals it embodies; all software should be free on principal, closed source is the devil, binary blobs are not exceptable, patents and copywrite should be abolished entirely.

The FSF does not want to abolish copywrites. They believe in them because with out it the GPL would not work.

#

If ignorance were a crime, Greene would be swinging from the gallows. His pathetically malinformed drivel

He probably knows he's talking nonsense. This is marketing-speak, not engineer-speak.

Reminder for those who've forgotten:

Q: How can you tell when a salesman is lying?

A: his lips move.

#

Across all IT organizations, what percentage of applications do you think that are used are open source? If open source is as superior as you claim, why would Roml Lefkowitz estimate that about 1-2% of IT applications are open source (or money spent on IT applications - I'm not sure which he meant)? I think it's because open source applications partially address IT needs,

It's because companies pushing proprietary software spend billions on advertising and marketing. They wouldn't spend all that money unless it worked.

#

"or money spent on IT applications - I'm not sure which he meant"

Of course you realize that anybody can implement a whole Open Source office environment (from servers to desktops) without paying a dime for software. So, this ambiguity makes it a pretty meaningless or even manipulative statistic, doesn't it?

"There have to be costs of using open source, from the expertise required to compile and run it, to the expertise required to understand what patches to apply."

I do not understand you, do you WANT there to be costs? Nobody ever said that Open Source is the magic bullet to solve all your IT demands. If one could implement an office software environment without paying money AND investing some effort, well... _smile_.

"Sometimes the UI is weak,... in Open Source" (italics mine) Oh yes please, lets talk user interface, and lets have a look at the Ipswitch products. Now there's some early 90's GUI crap if you ever saw it. And now have a look at ftp implementation in de KDE desktop, 100% transparent file management with integrated password management. And the links to servers can be put in your Bookmarks. (yes, Ipswitch does have a problem, and they know it)

So, sure one can find poor interface design in Open Source, but it's not unique to Open Source, at all.

So, what else is there to say here? Well, last year Ipswitch retracted their free ftp-light client (years ago everybody used it). I suppose that the company is very conscious of the problems they're facing with Open Source, and i'm sure that they know their products are, well... unremarkable compared to some Open Source stuff i know. My guess is that the company got itself a new bullish Chief to counter their problems. You see this stuff a lot with those proprietary software companies.

Jaap

#

My point about UI is not that there is inherently anything about it that makes it easier to do in a proprietary software environment (or harder in open source). It's that it's difficult to create good UI. It's also expensive, as developers who are good at it more often than not want to be paid for their work. Proprietary software companies have a financial model that allows them to invest in UI. They succeed to varying degrees, but if it's strategic and they focus on it, they can afford it.

This all comes back to a profit model for open source. The only ones I know of are to sell services (Red Hat) or sell hardware (appliance vendors) or sell advertising (Firefox - I think they are a foundation, but I believe they make a lot of money from Google's sponsorship).

That model doesn't work for all application categories. In those categories there are many current open source applications. Some of them are good. But on average their lack of a sustainable profit model means that adding all of the features that customers want, with a simple UI, doesn't happen, which is where proprietary software comes in, and why that market continues to grow.

Open source is not bad. Proprietary software is not good. They each have advantages. For small business IT and departments in enterprises, which is our market, we think an advertising model won't work, and service-based or hardware-based solutions are too expensive.

#

Which is why Hollywood is such a big user of Linux.. With a big budget and limited time they can't wait for someone to add a feature or fix a problem - they take the option of doing it themselves. From the story below is seems that they even require most of their commercial software suppliers to make their sources available.

A very interesting read:
<a href="http://digitalcontentproducer.com/dcc/revfeat/video_linux_hollywood/" title="digitalcon...oducer.com">http://digitalcontentproducer.com/dcc/revfeat/vid<nobr>e<wbr></nobr> o_linux_hollywood/</a digitalcon...oducer.com>

"At the software level, studios are using Linux versions of some of the leading commercial applications for 3D animation, compositing, special effects, and rendering — Alias Maya, Apple Shake, and Pixar RenderMan. Internally, the major studios have ported millions of lines of proprietary code to Linux and are creating their new programs in Linux."

"An irony of the migration of software to Linux is that Apple and Pixar became leading suppliers of Linux software. The most popular motion picture compositing software — Apple Shake (Linux in 2000) — and the most popular renderer — Pixar RenderMan (Linux in 1999) — are both sold by companies headed by Steve Jobs. Jobs hasn't made any pro-Linux statements regarding the future of his products (and recently Apple dropped the price of the Apple version of Shake so much that the Apple computer to run it seemed free). How the Mac/Linux equation will play out remains a concern for studios intent on controlling their destiny by staying with an open-source operating system rather than beholden to a proprietary third-party platform."

#

I saw one of Disney's animation studio offices full of cubicles a few years back at Walt Disney World, and sure enough, every computer was sporting the KDE 3.0 desktop. I'd recognize it anywhere. HOWEVER....



How ironic that Hollywood studios are using GNU/Linux and even contributing to some F/OSS projects (FilmGIMP, for example)...


<nobr> <wbr></nobr>...but those same Hollywood GNU/Linux-using studios don't want us to watch our legally-purchased DVDs on GNU/Linux! They try to sue us--even bring criminal charges against us--when we attempt to do so (e. g. "DVD Jon" Johansen). Might as well outlaw photocopiers or cameras....

#

I think we all know the difference between the craftsmen in the workshop and the artisans of the marketing department.

#

Come on Joe. I realize that an unfavorable article or statement about FOSS gets your idealistic panties in a wad but, your argument and your examples are terrible. You are also just as guilty of cherry picking the good projects while ignoring the many warts. It's different when you look at the big picture, warts and all.

Re: Myth #1

The "free as in speech, not as in beer" mantra that you quoted has to the most ridiculous mantra there is. You have to explain its meaning to everyone that hasn't heard it before because it isn't clear. Leading off your argument with this ambiguity is the beginning of your failure.

Yes, it is true that you can get reliable FOSS software off the net for free. But, we all know that there is still a cost incurred, even with Apache. The time and effort used to get the software working properly is expensive. Apache may be a shining example of the good in FOSS but, you know as well as I that there are many more abominations than there are Apaches. Even tools that are near perfect, but follow the Unix philosophy of many small job specific tools that get chained together to accomplish a task incur a time/labor cost. A good example of this is the Postfix/Amavis/SpamAssassin/ClamAV SMTP spam gateway. It is a great tool that works swimmingly, but it takes hours/days to setup and tune. Off the shelf proprietary software solutions that serve a similar role take only an hour or two to setup and configure in a similar fashion and they also have nice management interfaces and extensive reporting capabilities that the FOSS solution doesn't have.

No corporate IT, FOSS or proprietary is fee of labor costs.

Re: Myth #2

Again this is a two way street. While it is true that Apache and Firefox pump out patches daily, if needed, there are many other projects that don;t get this kind of response.

The kernel is a good example of this. Surely, even Joe has had a kernel update bork something important like say a SATA controller, and no resolution is made available for weeks/months. I remember a problem with the <a href="http://groups.google.com/groups/search?q=kvm+mouse+kernel&qt_s=Search+Groups" title="google.com">kernel KVM switches that went unfixed for years!!!!</a google.com> Proprietary software may not be the panacea that Greene suggests but, FOSS is just as guilty if not more so. But, you refuse to acknowledge ANY flaw.

Re Myth #3

You picked the worst possible examples for this myth. Have you ever used Nagios, GroundWork, Sendmail, et al in a corporate environment. Nagios takes days/weeks/months to get working properly in a corporate nevironment. IPSwitch What's Up Gold can do all that Nagios can, and a hell of a lot more, in under an hour!!!!!!!!!!!!

Finally, there is this statement:
That competition certainly divests him of an objective voice in speaking on the issue, since his bread and butter depends on besting those popular projects day in and day out, and provides him with a motive for slinging FUD as far and wide as he is able. Hello Pot? This is kettle.

Joe, you ought to look at yourself in the mirror, your objectivity is completely absent! Your "article" isn't a refutation of his, it is the tirade of an angry child. You lost credibility beginning at the title and continued to lose it throughout the article.

Greene was wrong about many things and his article was definitely self serving. But, Joe, your article is wrong about many things and if your article isn't self serving and pandering, then what is?

Best regards.

Your favorite "Microsoft Shill".

#

You must be a be on dialup for it to take so long. It couldn't be any easier to setup a fully configured antivirus/antispam email server with mailing lists from bare metal then by using Qmail Toaster Plus <a href="http://qtp.qmailtoaster.com/" title="qmailtoaster.com">http://qtp.qmailtoaster.com/</a qmailtoaster.com>. The support from the mailing list is more informative and helpful then I've seen in any other tech support.

I'm sorry that you perfer to use what you are told to use propiartarily. A shallow software user will miss many gems.

#

You almost had me convinced, but you just didn't use enough exclamation points. If only you'd added a few more, I'd have been totally sold on your point of view.

Well, now you know for next time. Generous portions of exclamation marks are the key to transforming opinion into fact!!!!!!!!!!!!!!!!1111111111111111!!!!!!!!!one!<nobr>!<wbr></nobr> !1111eleven!!

#

Once again, the MCSE rears its ugly head. I get a kick out of MCSE's like this poster, as he is clearly scared silly for his own job. I've seen plenty like that.

At my place of work, we *do* run Nagios on a rather large network. It is sweet, and it certainly didn't take "days/weeks/months" to get running right. It took me half a day...and virtually no maintenance afterwards. Oh, and we also run NMIS for history tracking. That took me less than an hour.

I don't use Sendmail, I personally prefer Postfix+Dovecot. But if you send email across the Internet, there's a 90% chance that it goes through some corporate Sendmail server. It took me about three hours to install, configure, *and* test my email server running Ubuntu GNU/Linux. So much for "days/weeks/months" to stand up a F/OSS server.

MCSE's don't know anything except point-n-click for MS Exchange and Microsoft Management Console. That's why this poster is so scared. When his MS Exchange Server tanks, he's got to call Microsoft to get him out of a jam for high dollar. We don't need to do that with F/OSS; we can--and do--fix it ourselves and keep stepping. That's what Freedom is about. Furthermore, F/OSS servers don't tend to tank nearly as often as those from Microsoft. And our data, unlike theirs, is in a freely-available file format. No vendor lock-in.

Have fun, little MCSE. You don't even know what freedom is anymore. When your MS Exchange Server corrupts your entire Information Store, thus putting you in the unemployment line, I'll be laughing my head off at you.

#

Hey now; He spent alot of money to become a certified MCSE surf. He can't go around allowing that expense to be invalidated by things like facts and historical evidencess

sssssahahahahahehe.. sorry, I couldn't get through it keeping a strait face.

#

I'm interested in this discussion but, I don't have the wherewithal to determine who is right. Could you provide some facts and history that might invalidate anything in the original post? So far, the only responses to the post have been name calling. As yet, no one has addressed any specifics regarding the post itself, or the facts therein. This would seem to reinforce the original poster's assertions and makes the name calling look like, well... name calling.

#

Good question, and one that I wish more folks would ask. Facts about this abound, and no, we don't mean Microsoft's "Get the Facts" ad campaign. Here's a little something to get you started.

Schools:
<a href="http://linuxplanet.com/linuxplanet/reports/4216/2/" title="linuxplanet.com">http://linuxplanet.com/linuxplanet/reports/4216/2<nobr>/<wbr></nobr> </a linuxplanet.com>
<a href="http://www.seattleschools.org/schools/hale/compu_pics/audit.htm" title="seattleschools.org">http://www.seattleschools.org/schools/hale/compu_<nobr>p<wbr></nobr> ics/audit.htm</a seattleschools.org>
<a href="http://archives.seul.org/seul/edu/May-2002/msg00129.html" title="seul.org">http://archives.seul.org/seul/edu/May-2002/msg001<nobr>2<wbr></nobr> 9.html</a seul.org>

Corporations:
<a href="http://news.com.com/2100-1023-212942.html" title="com.com">http://news.com.com/2100-1023-212942.html</a com.com>
<a href="http://news.com.com/2008-1082_3-5065859.html" title="com.com">http://news.com.com/2008-1082_3-5065859.html</a com.com>
<a href="http://blog.lobby4linux.com/index.php?/archives/83-No-One-Ever-Got-Fired-For-Using-Microsoft.-Yes-They-Did..html" title="lobby4linux.com">http://blog.lobby4linux.com/index.php?/archives/8<nobr>3<wbr></nobr> -No-One-Ever-Got-Fired-For-Using-Microsoft.-Yes-T<nobr>h<wbr></nobr> ey-Did..html</a lobby4linux.com>
<a href="http://blog.lobby4linux.com/index.php?/archives/85-One-Small-Business-Gladly-Gives-Microsoft-the-Boot.html" title="lobby4linux.com">http://blog.lobby4linux.com/index.php?/archives/8<nobr>5<wbr></nobr> -One-Small-Business-Gladly-Gives-Microsoft-the-Bo<nobr>o<wbr></nobr> t.html</a lobby4linux.com>

Government:
<a href="http://www.largo.com/index.cfm?action=dept&drill=it" title="largo.com">http://www.largo.com/index.cfm?action=dept&drill=<nobr>i<wbr></nobr> t</a largo.com>
<a href="http://business.newsforge.com/article.pl?sid=02/12/04/2346215&tid=37&tid=110&tid=2&tid=132&tid=3&tid=31" title="newsforge.com">http://business.newsforge.com/article.pl?sid=02/1<nobr>2<wbr></nobr> /04/2346215&tid=37&tid=110&tid=2&tid=132&tid=3&ti<nobr>d<wbr></nobr> =31</a newsforge.com>

There is much, much more out there.

#

"Come on Joe. I realize that an unfavorable article or statement about FOSS gets your idealistic panties in a wad but, your argument and your examples are terrible. You are also just as guilty of cherry picking the good projects while ignoring the many warts. It's different when you look at the big picture, warts and all."

If you didn't lie so much, people might not always be mistaking you for a Microsoft shill.

#

"If you didn't lie so much, people might not always be mistaking you for a Microsoft shill."

Uh, Joe? Are you sure that's really a mistake?

;)

#

Quote: 'The "free as in speech, not as in beer" mantra that you quoted has to the most ridiculous mantra there is. You have to explain its meaning to everyone that hasn't heard it before because it isn't clear. Leading off your argument with this ambiguity is the beginning of your failure.'

The real ambiguity (for which the speech/beer analogy was created) lies in the term 'Free':

<a href="http://en.wiktionary.org/wiki/free" title="wiktionary.org">http://en.wiktionary.org/wiki/free</a wiktionary.org> <a href="http://www.google.com/search?&q=define%3A+free" title="google.com">http://www.google.com/search?&q=define%3A+free</a google.com>

Beer is a commonly understandable commodity, thusly it serves well to illustrate the difference between the very prevalent (and capitalistically distorted) version "FREE!*" and the more general 'free as in Liberty.'

IMO when one explicitly phrases the mantra along the lines of "Free Software, means free as in 'free speech', not as in 'free beer'" it isn't altogether confusing.

Meow, if I were to go around saying "..free as in free speech, not as in free Ochre" I would leave more than a few people perplexed and in need of clarification: not the color [#CC7722] but <a href="http://www.ochremusic.com/music" title="ochremusic.com">http://www.ochremusic.com/music</a ochremusic.com> (no ogg but he does use LAME.)

Follow?

Ok, back on topic. The GNU Project has some advice :
<a href="http://www.gnu.org/philosophy/words-to-avoid.html#ForFree" title="gnu.org">http://www.gnu.org/philosophy/words-to-avoid.html<nobr>#<wbr></nobr> ForFree</a gnu.org>

Lastly, here is a supporting Wikipedia article:
<a href="http://en.wikipedia.org/wiki/Gratis_versus_Libre" title="wikipedia.org">http://en.wikipedia.org/wiki/Gratis_versus_Libre</a wikipedia.org>

Thanks for your time and now lets effectively explain and promote Libre Software!<nobr> <wbr></nobr>:-D

#

Best go back to the drawing board, Mr Greene. You obviously have not done any reserach - if you have then you lack comprehension.

#

Address the point: In spite of your assertions, why is it that the FOSS world has produced UIs superior to your own products?

#

Joe, who are you kidding?

Organizations on a tight budget simply can't afford to pay all sorts of license fees for software. For even a small enterprise, these fees mount to 100s of 1000s of dollars.

As for your other assertions, paid-for-apps have t he worst documentation in the industry. Microsoft and Oracle have crappy support. Just search Google and see how users are forced to support themselves. Free software has embraced self-support and it works very well, execpt you don't have to pay cash for it.

Free software is filling a need that you can't. Your article is nothing but sour grapes. As the collective knowledge contained in all that free software increases with time, there will be less and less need for proprietary software companies like yours.

Get real. The rest of us are.

#

Sorry, I meant "Roger, who are you kidding..." not "Joe, who are you kidding" in my last post.

#

...open mouth wider and insert other foot.

IIS has larger market share???

Not only is your recollection faulty, your intent and integrity is questionable when publishing an article with no better research than "my recollection is".

#

<a href="http://www.ducea.com/2006/06/09/netcraft-web-server-survey-or-should-i-say-domain-parking-survey/" title="ducea.com">The largest movement of sites from Apache to IIS was once again at Go Daddy, with over 1.6M hostnames moving from Apache to IIS this month. </a ducea.com>


sometimes ms pays some of the parking domain people money to have the netcraft stats look a little better.

#

That's interesting. What do you think of that Roger?

#

Roger<nobr> <wbr></nobr>... if what I'm hearing from you is true, your presumption is that ALL proprietary software is devoid of warts, moles, festering sores, and downright abominations of applications? Because I've used plenty of proprietary apps in my life, and while F/OSS apps may not all be perfect, they fare a whole lot better than their proprietary cousins.

#

As of v8.15 of Imail anyway - they're using OpenLDAP for their integrated LDAP service. Hmmmm.

#

If you didn't lie so much, people might not always be mistaking you for a Microsoft shill.

There's that dull wit again Joe. It's no lie. You seem to look at the world through rose colored where all FOSS is wonderful and all proprietary software/Microsoft is bad.

I commend you for being passionate about FOSS, everyone should have a passion but, you shouldn't let that passion blind you to the facts. When you do that, you become just another zealot. Oops, too late. But, it's never too late to improve yourself, Joe. Next time try a little less passion and a little more fact.

Your favorite "Microsoft shill".

#

Perhaps you'd prefer Microsoft's "Get the Facts", eh? Or would you prefer *actual* fact?

If it's the latter, then here are some facts:

<a href="http://www.k12ltsp.org/press_freedom_day.html" title="k12ltsp.org">http://www.k12ltsp.org/press_freedom_day.html</a k12ltsp.org>
<a href="http://news.bbc.co.uk/2/hi/europe/6332441.stm" title="bbc.co.uk">http://news.bbc.co.uk/2/hi/europe/6332441.stm</a bbc.co.uk>
<a href="http://www.neowin.net/forum/index.php?showtopic=452150&mode=threaded" title="neowin.net">http://www.neowin.net/forum/index.php?showtopic=4<nobr>5<wbr></nobr> 2150&mode=threaded</a neowin.net>

And if you wonder which is the most popular Web server on the Internet...it ain't Microsoft IIS.

<a href="http://news.netcraft.com/" title="netcraft.com">http://news.netcraft.com/</a netcraft.com>

The root DNS servers aren't powered by Microsoft DNS, either. They typically run BIND on some kind of UNIX/Linux/BSD. Why? Because they don't fall over under heavy load.

Now, refute these facts, little MCSE, and improve *yourself.*

#

At no point did I suggest that Unix or Linux based software wasn't widely used. Never have I denied that Linux is being used with good results in some places.

What I clearly stated was that FOSS is not a panacea. I also stated that while Joe rants on about how all proprietary software is inferior and says that all FOSS is superior, even if only because it is free, the reality is very different. Reread the posts. They say that some FOSS is excellent and some is utter crap. It is the FACT that some is utter crap that Joe, and now you seem unwilling to accept.

Your links do not refute the fact that some/much of FOSS is utter crap! While Greene and Joe were on about Apache, my only reference to it was that it was excellent and that it was the exception in the FOSS world rather than the norm. But your misunderstanding is your problem, not mine.

#

You might want to reread the end of the article. Joe did not say "all proprietary software is inferior and...all FOSS is superior".

Here's the start of Joe's last paragraph: "There are proprietary software products that are better than their open source equivalents. There are critical software products that have no open source equivalent. There are open source projects of dubious quality and reliability. Network World readers learned none of that from this Face-off."

Perhaps you missed that. Joe might not be a bastion of objectivity, but I appreciate the tenor of his response.

- T

#

You don't even know what "Free" means. You state your own confusion of it. Your language is that of Steve Ballmer.

Your posts are so full of invective and distortions that I now believe that either you're a Microsoft employee or Rob Enderle. You have not addressed *any* of the facts that I presented, and you won't, because you can't.

Your "presentation" does nothing to show how I have more freedom with F/OSS than with proprietary software. It does not show me how I am any less subject to BSA-style raids (or the threat thereof) against my company or school district with proprietary software. It doesn't show me how I avoid vendor lock-in with proprietary software. Instead, you have shown nothing except your bias for proprietary software, and Microsoft's in particular.

If you're so in love with Microsoft, then by all means, continue to use their stuff. Good luck, kiddo! I'll stick with freedom, and that means F/OSS.

#

While Greene and Joe were on about Apache, my only reference to it was that it was excellent and that it was the exception in the FOSS world rather than the norm.

If Apache was the exception, then give some examples of the norm.

If we take Apache as the exception, as you say, then we also have to include software of similar quality as exceptions also. Take, for example, the 19,000 or so packages available for Debian Etch. Let's presume that there's more than just Apache as an exception in there (otherwise, all our Debian machines would constantly have problems), so let's say that there are 9,000 unstable packages - that is, packages that are not of similar quality to Apache. I would strongly disagree with this, being a Debian user, but let's presume for arguments sake. That says there are 10,000 software packages of similar quality to Apache in just the Debian Etch package archive.

Now, I don't know what it would take for something to be considered an exception - would it be 1 in 10, 1 in 100, 1 in 1,000? Any way you look at it, that means there must be (in your opinion of course) hundreds of thousands of F/OSS software packages out there that are not fit for production use.

In my experience (over 7 years now) with GNU/Linux and F/OSS software, I've found it to be more stable, reliable and intuitive than any proprietary software I've used. If you really think that unstable and unuseable F/OSS software is the norm, then you should try using some of it. This is simply not true.

#

The only way to be sure is to have and use the source for the entire toolchain:
<a href="http://www.acm.org/classics/sep95/" title="acm.org">Reflections on Trusting Trust</a acm.org>. Ken Thompson. Communication of the ACM, Vol. 27, No.
8, August 1984, pp. 761-763.

#

"Greene is CEO of Ipswitch, a leading developer of network monitoring, messaging and secure managed file-transfer software."

do not use open source and become a knowledgeable user - keep buying my software and do what we say.

#

Can't really speak for the companies here, but I've been using Open Source Software almost exclusively for the last few years for all my home/college work(except for gaming).

Sure, there are some programs that make me want to cry when looking at the GUI, but then again a GUI isn't the only UI a program has. Nowadays I find myself increasingly comfortable with CLI applications.

I don't want this to be a GUI VS CLI flamewar, but for my own purposes: CLI is pure gold.

"developers who are good at it[Designing good [G]UI] more often than not want to be paid for their work."
Isn't that true for pretty much any kind of software?
People who are good at device driver development would like to be payed for it.
People who are good at writing libraries would like to be payed for it.
Yet we do see great drivers and libraries in the OSS landscape. Why are GUIs any different?

#

I believe GUIs are different for three reasons. Let's contrast GUI development with device driver development. First, I would argue that creating good, usable GUIs requires rather different skills than general development, while writing quality device drivers requires skills similar to general development, but at an arguably higher level of capability.

Second, I believe there is a perception problem. Writing good device drivers is considered more "worthy" than building a good GUI. It's easy to imagine an OSS developer receiving respect for device driver (or kernel, and so on) development. Now imagine the response to "Hey everyone, I just finished this GUI - look how usable it is!" I think we all know there's a difference in the perceived value to the community. I'm not trying to say that OSS is hostile to good GUIs, or that there are no OSS developers interested in creating quality GUIs. However, the general appreciation is lower, and I believe this perception leads to lower interest in GUI development, with a resulting deficit in good, usable GUIs in OSS.

The third reason: the CLI is culturally more accepted than the GUI in OSS. I don't think that is in any way a problem by itself - it's just a difference. But I believe it further explains the lack of interest in GUIs among OSS developers.

- T

#

Example:
The<nobr> <wbr></nobr>.Net hype created quite a few ASP.NET hosts, but when the hype died away so did a great deal of those new IIS servers.

Result: Fluctuation

#

Makes sense, but what then explains IIS gaining market share more recently? Maybe they're getting some traction with<nobr> <wbr></nobr>.Net?

#

To each his own.

Some problems are better solved using a piece of software that is not OSS, but don't go around telling people all OSS is bad.

The difference between OSS software and your kind of software is that OSS continues to evolve and draw strength from each piece of code that's written. This is something only a company with enormous coverage(like microsoft) could pull off, and all thanks to the GPL.

So while most companies have to start from scratch(or close enough) OSS projects(and companies that embrace it) can start where all the others have worked to get. Pretty similar to the way we teach children: each child learns whatever his ancestors have learned and adds his/her own bit later on.

All that leads me to think that the OSS model has an exponential curve in terms of quality(or at least lines of code being written) that is why (IMNSHO) it'll conquer the [software] world sooner or later.

#

given Joe's erudite analysis...

For security fixes only maniacal Unix geeks compile their own software; the rest of us use apt:

apt-get update; apt-get -yqq upgrade

See? Is Microsoft that responsive? I mean, really!

Geek Unorthodox

Relax, have a homebrew!

#

Before IE bundling drove Netscape completely bankrupt, the CEO (after reading Cathedral and the Bazaar) worked with Eric Raymond on a new business strategy resulting in Netscape becoming The Mozilla Foundation and releaseing Netscape's source code as the Mozilla browser. They now lead the Firefox browser and Thunderbird email program projects.

They also registered Open Source as a foundation and started Opensource.org where Mr Raymond is still chairman of the board (I believe).

#

That's what happens when Microsoft pays folks like GoDaddy to move parked (i. e. reserved, but not actually being used for something) domains over to some IIS servers.

The real question is, why has Apache been *so* dominant for ten years straight and continuing? Having used both Apache and IIS, my answer is that Apache kicks IIS in the teeth all the way around--performance, security, reliability, cost to maintain, freedom from lawsuits--in short, just about everything.

We're a Microsoft shop...but our Internet Web servers all run Apache on either Red Hat Enterprise Linux (the increasing majority) or Solaris. Even our Windows- and IIS-loving server manager (a die-hard MCSE) won't trust Windows/IIS to the Internet. Now, *that* is pretty telling.

We also run Nagios and NMIS, on RHEL and Debian respectively. They replaced HP OpenView NNM. There's no way in *HELL* we will go back to HPOV NNM or anything else proprietary; the F/OSS solutions are just too damned good. And if I need to, I can tweak it to our particular needs; can't do *that* with Microsoft.

So, given our results, I will continue to trust our network to F/OSS, and increasingly so. I *love* it.

#

FileZilla is an awsome FTP client. I use to use a shareware product that I believe was called wsFTP. Don't know what happened to that FTP client, I gave up using it years ago. I think FileZilla has a much better user interface. Can't beat the fact that there are no pop up screens or limitations. I was so impressed I began doing something I never use to do, I donated to the project via pay-pal.
FileZilla User

#

I use to use a shareware product that I believe was called wsFTP. Don't know what happened to that FTP client

You were using wsFTP Lite from Ipswitch. Robert Greene, who this debacle is all centered around is the CEO of Ipswitch.

You stopped using wsFTP for two reasons. First, Microsoft started including graphical FTP capability in Internet Explorer, which significantly reduced the need for a separate FTP client. Second, most downloads changed from FTP to HTTP, which further reduced the need for FTP clients.

If you're nostalgic for wsFTP but use Linux, you can get a wsFTP Lite wannabe in <a href="http://gftp.seul.org/screenshots.html" title="seul.org">gFTP</a seul.org>. Filezilla is much more advanced and feature rich than gFTP but, it is also much much fatter.

#

You're right...that is rather funny. Back in my Windows-using days, I, too, used WS-FTP. At the time, it was the best thing going for Windows users that I had encountered. I now steer newbies to using gFTP (I myself use wget).

And that is the kind of freedom that you have with F/OSS platforms. You have much more choice on what goes onto your computer than with Microsoft or Apple. I wouldn't put it past Microsoft to make Vista disallow the installation of certain apps that MS doesn't like (e. g. OpenOffice.org, Firefox, or even maybe WS-FTP). You can't do that with F/OSS.

#

More misconceptions:

1) F/OSS developers don't get paid for their work.

Fact: Novell employs lots of the Gnome GUI desktop developers. Red Hat employs lots of Linux kernel developers. They get paid. Lots of hardware manufacturers pay their developers to write both Windows and Linux hardware drivers. Most OSS developers get paid a salary to do so.

2) Most 98-99% of IT applications are proprietary, not F/OSS.

Fact: You're leaving out most infrastructure software: http servers, browsers, DNS servers (how much of the Internet depends on Active Directory? Almost zero. bind is king).

What about network appliances like firewalls, load balancers, proxy gateways? Name one - just one - firewall appliance vendor who bases their product on Microsoft ISA. Virtually all are NetBSD, OpenBSD or Linux-based.

Good luck, Mr. Greene, turning Ipswitch into a winner. I suggest you go back and do your homework before you make public statements. Can you imagine how your employees feel seeing you embarrassed in public like this?

#

3) Poor users of F/OSS have to compile source code on their own for the majority of their applications. And they have to compile in the patches as well.

Again, you're implying that this is the truth. It hasn't been true for years. Red Hat's RPM and Debian's DEB package formats are by far the norm. Any Linux distro* worth it's salt has an online repository where users can (automatically in most cases) get updates and new software in binary format.

Of course, there's always the wicked-cool Gentoo for those that want a Linux install completely optimized for your particular hardware setup. Compiled from source - pretty easily, too!

#

One of my points was that a large number of open source developers work for companies that are paying them. Those companies make money, and they often do so by selling services. They typically don't by make money by selling software. Customers who pay Red Hat and Novell for services are not doing so for a few thousand dollars - it adds up quickly. Shrinkwrap, proprietary software from vendors who make money principally from selling the licenses is in many cases much less expensive than the service model.

As for the 98-99% comment, I was referencing a speech by r0ml Lefkowitz, who I think has an interesting perspective on open source. I'll keep searching for the podcast; I can't find it now. It would be more productive to quote exactly what he said. I did, however, find another podcast of his, in which he talks about what it takes for an open source project to take root. I think his list is missing something about profit, but is an interesting list. The rest of this post is from the following link.

<a href="http://www.itconversations.com/shows/detail139.html" title="itconversations.com">http://www.itconversations.com/shows/detail139.ht<nobr>m<wbr></nobr> l</a itconversations.com>

For an Open Source project to take root, it must meet several criteria. R0ml proposes that these are:

a) Everybody (for some definition of everybody) needs it or can use it.
b) Those who use it (everybody, by definition above) want to improve it (for some definition of improve).
c) Significant business value associated with the use of the software does not reside in the software itself.
d) It's cool.

R0ml examines four major focus areas within the IT department at AT&T Wireless with respect to these criteria, and explores reasons why there aren't any credible open source projects around these; whether there should (or should not) be; and what they might look like.

[Full title: "(More) Missing Open Source Projects." This keynote presentation was recorded at the Open Source Business Conference 2004 held in San Francisco, CA. A PDF file of this presentation's graphics is available. See the complete list of OSBC 2004 presentations on IT Conversations.]

Robert M. "r0ml" Lefkowitz is the Chief Technical Architect and Vice President of Data Services for Information Technology at AT&T Wireless.

#

Regarding the second point, you reinforce one of my points, which is that open source works best when there's a profit model such as service or hardware or additional proprietary software or all of the above that goes along with it. Open source makes good sense as a component of a product, but, although there are exceptions, not as an entire product. The economics don't work for many applications. If they did, there wouldn't be a commercial software market. I don't see that changing.

#

``But I stand by my core assertion that for many applications, it makes more sense for organizations with scarce resources and tight budgets to purchase software from a vendor they hold accountable for keeping the software up to date.''

But the vendor merely supplies the patches. Maybe. I spend way too much time downloading patches from vendor's web sites and applying them. (Hmm... seems pretty much like what I have to do for OSS packages.) So for $20K a year in support costs -- and that's just an amount that comes to mind for a single package we use at work -- I am blessed with the privilege of having access to their web site. I still have to download patches and apply them myself. Automated patching? Naw, that's too hard. Move the old files aside and copy news onto the system yourself. Maybe if we spent $50K/year or $100K/year... Crimeny, I even have to download the documentation myself. They can't even be bothered to cut a CD and ship it to me as part of the support. And don't get me started on the "quality" (if one can call it that) of the phone support I receive from the vendor.

For the OSS packages I support, I get the software for nothing. I get support from the package's web site where the forums typically are monitored by some of the core developers. I can make suggestions and they sometime make it into the code. Response time? It might take a day or two. Similar to the level of response I get from commercial software vendors once we play phone tag.

The vendors of the commercial packages typically want us to pay a lot for for support if we actually want to be allowed to make suggestions for the software that would help fill our needs. They'll listen for all that extra money. One of the critical commercial packages I install and support hasn't changed noticeably in years. This despite being told by their sales folks that they actually have customers who pay all that extra money for input into the development process. I frankly haven't seen any change in the software's appearance or performance in something like eight years. On the other hand, some of the OSS projects whose software I use incorporate change so quickly that I sometimes find it hard to keep up. Of course, you'll probably now say something to the effect of "But corporations need that stability blah blah blah...". Hogwash. We're all sick and tired of inferior software dressed up in flashy packaging that doesn't perform, doesn't get any better from release to release, and costs us an arm and a leg.

There's a reaon why many organizations are strapped for cash. They blew it on expensive proprietary, commercial software packages. They didn't get much benefit for all that expense, either. The IT people still have to slave over keeping the proprietary package current and up and running. If I have to put the effort into keeping software current and running, doesn't it make sense to use the software that doesn't require all the cash outlays?

#

There are too many examples of commercial software packages that don't provide enough value, whether that's through lack of adding needed features, sub-par UI, poor service, or weak documentation. But the fact that these products even exist says something about how open source isn't a complete solution for many product categories. It's also an opportunity for other companies to do it better, which is what makes for a reasonably dynamic market with new vendors displacing ones who don't keep up. Unfortunately for customers, that survival of the fittest process takes a while, and for stretches of time there may be no products, either open source or commercial, that do what they need and do it well.

When commercial software works well, it adds real value through the combination of functionality, UI, documentation, support and updates. Our business model requires that we make money from selling licenses, and that support costs are low. When we do that well, few people call for support because for most customers we have fully addressed their problem with a product that is reasonably easy to learn and use. When we don't do it so well, as has happened at times, more customers call, our support costs rise, and our profits decline. Long term, the only way out of that spiral is to improve the quality of our software so people get what they need without having to call support. That's a powerful motivator. With open source, the incentive isn't as great to produce a complete product. Yes, there are examples, but there are many more areas where open source technology provides part of the solution, but not all of it. That's beneficial to vendors, who, for instance, through using open source libraries are able to reduce development costs and get products to market faster. In markets with a number of competitors, users benefit from these cost savings through lower license fees.

On a separate note, if you (and I pose this to others as well) were running a company with a hundred to a few hundred employees, what percentage of IT applications that you used would be completely open source?

#

> But my recollection is that IIS has larger market share, and the gap is widening.

This was the point when I fell out of my chair laughing...

#

> it makes more sense for organizations with scarce resources and tight budgets to purchase software from a vendor they hold accountable for keeping the software up to date.

There are lots of things that "make sense", but still aren't true. It's the old "back of envelope calculation" fallacy. Start with bad assumptions, and you get bad results, no matter how logical the steps are in between. What evidence do you have that proprietary products are more "up to date" than their FOSS counterparts? How about the example of IE? Microsoft's crown software product, the one they fostered specifically to try to lock the entire Internet into their server products, lay essentially dormant (aside from weekly security updates), for a half dozen years or so, until Firefox came along and knocked them out of their stupor. Microsoft still has a substantial lead in market share with IE, by reason of sheer momentum, but they are still losing ground at a steady pace. (FF is up around 25% by most calculations.)

I'm not sure what planet you live on to make you believe that proprietary software vendors, especially those with a monopoly, or near monopoly in one area or another, are anything approaching "responsive" to their customers. Take a wander over to Ed Foster's <a href="http://www.gripe2ed.com/" title="gripe2ed.com">"Gripe Line"</a gripe2ed.com> if you want a shining example of great customer support from all of your proprietary friends.

#

I agree that as a market approaches a monopoly there is less incentive for a vendor to invest in innovation. Most software markets aren't monopolies, though, and especially as you move away from utilities and development tools (which matter most to those writing open source), there is more incentive for commercial vendors to maintain and enhance applications than there is among the open source community.

As for IE losing ground to Firefox, I suspect that Microsoft hasn't invested more in IE because they don't view Firefox as as much of a threat as they did Netscape, and they also have a lot of other priorities. I think a big part of Firefox's success is that they have a lot of revenue from Google to invest in development, which is not the case for most open source projects. I think it's significantly because of their revenue that they have built such a nice application.

#

Roger, please stop postulating about anything related to open source vs. closed source. You clearly haven't been following it for long enough to have a good understanding of how it works, how IT makes decisions about it, and how it is related to, competes with, integrates with, and replaces closed source/commercial software. You can trivialize these things all you want, but it doesn't make them trivial.

There's no traction to be had with<nobr> <wbr></nobr>.Net. It's been long enough now that if something significant happened with<nobr> <wbr></nobr>.Net, it would be called a "comeback". There's absolutely no reason to believe that this is taking place. A single chart is only good for making really broad statements like "Apache has dominated the web server space for years". It's *not* a good source for speculation as to why the chart itself shows fluctuations.

#

Why do you think market share fluctuates?

#

"But my recollection is that IIS has larger market share, and the gap is widening."

Sorry, Roger, you are dead wrong. Check this <a href="http://survey.netcraft.com/Reports/current/graphs.html" title="netcraft.com">Netcraft chart</a netcraft.com> and you'll see that IIS lags badly, and has for a number of years. The gap is widening, but it is in the other direction as Apache continues to dominate.

By the way, Why do you have to reboot your Win2000 server every week? Can't a proprietary server stay up for more than 6 days? That is pathetic.

#

You really should look at <a href="http://news.netcraft.com/archives/2007/02/23/march_2007_web_server_survey.html" title="netcraft.com">this page</a netcraft.com> for a recent Netcraft survey. Some of the links on the Netcraft site are out of date or 404.


Apache still beats Microsoft by almost 2-to-1, but sometimes the gap narrows.

#

That's interesting. Why do you think market share between Apache and IIS fluctuates so much over time?

#

Thanks for the data on IIS vs. Apache. It looks like Microsoft was catching up a few years ago and now Apache is gaining again, and as you say, is in the lead.

Even if I had been right on my recollection, Apache is an example of a successful open source application. There are others. But I stand by my core assertion that for many applications, it makes more sense for organizations with scarce resources and tight budgets to purchase software from a vendor they hold accountable for keeping the software up to date.

IT departments spend money on proprietary applications when they believe it's in their best interests to do so. More often than not, that's the decision they make.

#

Greene's firm hawks software for network monitoring, secure file transfers, and email servers. Let's see. Could the competition from open source products like Nagios and GroundWork Open Source; FileZilla; and Sendmail, Qmail, and Postfix give him an axe to grind?



Joe! Good grief! How could you miss <a href="http://www.opennms.org/" title="opennms.org">OpenNMS</a opennms.org> in that list. I mean, it's not like Tarus Balog, the OpenNMS maintainer was the <a href="http://www.networkworld.com/columnists/2007/030507-faceoff-open-source-yes.html" title="networkworld.com">counterpoint column in the "face-off"</a networkworld.com>. Oh wait! In fact, he was!

So, Joe, please, give credit where credit is due. OpenNMS will certainly give IpSwitch's network monitoring products a run for their money. And, <a href="http://searchnetworking.techtarget.com/productsOfTheYearCategory/0,294802,sid7_tax306257_ayr2007,00.html" title="techtarget.com">I'm not the only one that thinks so</a techtarget.com>.

#

"Can you really afford to wait for a [vendor] to agree with you on the urgency of action if your network is down?"



There. Fixed it for him.<nobr> <wbr></nobr>;-)

#

I think it's because open source applications partially address IT needs, for certain kinds of customers



Proprietary products tend to address IT needs only partially, and this on a consistent basis. When open source software (OSS) is selected, the chances of complete needs-met are much higher. After all, a vendor is looking to hit the market in it's sweet spot (best bang for buck). OSS is almost always developed by IT people for IT people.



There have to be costs of using open source, from the expertise required to compile and run it, to the expertise required to understand what patches to apply.



Actually the costs associated with deploying proprietary IT infrastructure are much higher -- even if one simply levels the playing field and says "we will buy only free, downloadable software" in each case. With OSS, deployment is faster (time is money), more reliable (fewer bugs), by its very nature, community based (service) and also more robust (scalable). Therefore, it's cheaper, period.



Sometimes the UI is weak, sometimes the documentation is incomplete.



The UI may not be "intuitive" (especially to a lay person). But it is hardly ever "weak". Would you prefer going to a Dr. who merely clicked around on some GUI interface to "discover" your health status? Or one that was a bit more hands-on?



Believe it or not, the documentation for OSS products is rarely ever "incomplete". Nearly any answer to any question can be found, eventually, using Google. On the other hand, even when a user community forms around a proprietary package, users are often left with a lot of guesswork (to see inside the black box that is closed source).



As for greed, are you saying that open source proponents are principally altruistic?



Actually the answer is yes and no. It would appear that most of us have day jobs that *pay* us to work on OSS. So, just imagine dozens of paid programmers, none of whom work for the same company, working together to solve a problem. Are we being altruistic? Yes. Are we doing it for the money? Indirectly, yes.



A proprietary vendor creates a package to make a buck. Pretty much end-of-commitment. OSS developers develop software to save a buck. We do so collaboratively, not just to spread the wealth, but to also spread the burden. A continuous commitment.



Guess which "model" will most likely win?<nobr> <wbr></nobr>;-) Methinks most OSS providers aren't all that worried about making a buck on the service end of OSS products, either. It isn't about profit, it's often not even about savings. Sometimes it's just nice to work on something that's incredible and "just works".

#

Hey, Joe didn't say that nobody in the OSS world is greedy. He just clearly stated thay you are nowhere close to being a disinterested party in this dialog, and when you claim that OSS is bad that coincidentially this pulls the blanket your way.

This was a very clear line in his article, why do you have to put words in this mouth?

#

Joe,

Across all IT organizations, what percentage of applications do you think that are used are open source?

If open source is as superior as you claim, why would Roml Lefkowitz estimate that about 1-2% of IT applications are open source (or money spent on IT applications - I'm not sure which he meant)?

I think it's because open source applications partially address IT needs, for certain kinds of customers. You talk about people downloading and compiling applications from the net. What percent of companies want to spend their IT resources doing that, and what does it cost?

In some cases, like Apache, it makes sense for companies to allocate resources that way. But my recollection is that IIS has larger market share, and the gap is widening. Why would that be if open source were compelling for all users?

There have to be costs of using open source, from the expertise required to compile and run it, to the expertise required to understand what patches to apply. Sometimes the UI is weak, sometimes the documentation is incomplete. There are cases where open source applications have excellent UI and documentation. But that's hard to do, and there is only the motivation to do that in selected areas.

As for greed, are you saying that open source proponents are principally altruistic? My impression is that open source companies use that model because they think it will make more money, and that they plan to do that by charging for services, or perhaps hardware. I don't see why that model would appeal to customers with small IT budgets.

#